performance of drained and undrained flexible pavement ... report.pdfthe 100-kn load present in the...
TRANSCRIPT
Performance of Drained and Undrained Flexible Pavement
Structures under Wet Conditions
Test Data from Accelerated Pavement Test Section 543-Drained
Draft Report Prepared for California Department of Transportation
By:
Manuel O. Bejarano, John T. Harvey, Abdikarim Ali, Mark Russo, David Mahama, Dave Hung, and Pitipat Preedonant
February 2004
Pavement Research Center Institute of Transportation Studies University of California at Berkeley
TABLE OF CONTENTS
Table of Contents............................................................................................................................. i
List of Figures ................................................................................................................................. v
List of Tables ................................................................................................................................. ix
Executive Summary ....................................................................................................................... xi
1.0 Introduction......................................................................................................................... 1
1.1 Objectives ....................................................................................................................... 2
1.2 Test Sequence ................................................................................................................. 2
1.3 Organization of Report ................................................................................................... 3
2.0 Test Program....................................................................................................................... 5
2.1 Test Section Layout ........................................................................................................ 5
2.2 Environmental Conditions .............................................................................................. 6
2.3 Instrumentation ............................................................................................................... 6
2.4 Test Program................................................................................................................... 8
2.4.1 Falling Weight Deflectometer Testing........................................................................ 8
2.4.2 Heavy Vehicle Simulator Trafficking....................................................................... 10
2.4.3 Data Collection ......................................................................................................... 10
3.0 Summary of Test Data ...................................................................................................... 13
3.1 Temperature and Moisture Condition Data .................................................................. 13
3.1.1 Temperature Data...................................................................................................... 13
3.1.2 Rainfall and Moisture Content Data ......................................................................... 14
3.2 Permanent Deformation ................................................................................................ 18
3.2.1 Surface Rutting Measured with the Laser Profilometer ........................................... 19
3.2.2 In-depth Permanent Deformation ............................................................................. 21
i
3.3 Comparisons of the Performance of Similar Test Sections with Dry Base Conditions 21
3.4 Elastic Deflections ........................................................................................................ 24
3.4.1 Surface Elastic Deflection Data ................................................................................ 24
3.4.2 In-depth Elastic Deflections...................................................................................... 25
3.4.3 Back-calculated Moduli from In-depth Elastic Deflections ..................................... 30
3.4.4 Comparison with Previous Section Tested under Dry Conditions ........................... 33
3.5 Surface Cracking........................................................................................................... 33
3.6 FWD Test Results ......................................................................................................... 35
3.6.1 Deflections ................................................................................................................ 37
3.6.2 Back-calculated Moduli from FWD Deflections...................................................... 38
3.7 Ground Penetrating Radar (GPR) Surveys ................................................................... 43
3.8 Forensic Activities ........................................................................................................ 46
3.8.1 Extracted Cores......................................................................................................... 46
3.8.2 Percolation Tests....................................................................................................... 52
3.8.3 Air-Void Content ...................................................................................................... 56
3.8.4 Dynamic Cone Penetrometer (DCP) Data ................................................................ 57
3.8.5 Trench Data............................................................................................................... 60
4.0 Performance Evaluation and Mechanistic Analysis ......................................................... 65
4.1 Pavement Responses..................................................................................................... 65
4.2 Fatigue Analysis............................................................................................................ 67
4.3 Economic Analysis ....................................................................................................... 69
4.4 Rutting Analysis............................................................................................................ 70
5.0 Summary and Conclusions ............................................................................................... 73
ii
5.1 Conclusions................................................................................................................... 74
5.2 Recommendations......................................................................................................... 75
6.0 References......................................................................................................................... 77
iii
iv
LIST OF FIGURES
Figure 1. Section 543 layout and locations of MDD and RSD measurements............................... 5
Figure 2. Depths of MDDs, thermocouples, and water infiltration holes in Section 543............... 7
Figure 3. Schedule for test program................................................................................................ 9
Figure 4. Average air temperatures inside and outside HVS temperature control box,
Section 543............................................................................................................................ 15
Figure 5. Pavement temperatures at depth, Section 543............................................................... 15
Figure 6. Average monthly rainfall during testing of Section 543. ............................................. 16
Figure 7. Volumetric moisture content in agregate base and subbase layers. .............................. 16
Figure 8. Volumetric moisture content in the subgrade at various depths from the surface. ....... 17
Figure 9. Moisture contents from aggregate samples extracted during forensic study. ............... 18
Figure 10. Average maximum rut depth for Section 543. ............................................................ 19
Figure 11. Section 543 surface rut profiles at various stages of HVS trafficking. ....................... 20
Figure 12. In-depth permanent deformation in Section 543. ........................................................ 22
Figure 13. Contribution of pavement layers to surface rutting in Section 543............................. 22
Figure 14. In-depth permanent deformation from Goal 1/Goal 3 test sections. ........................... 24
Figure 15. RSD deflections for Section 543. ................................................................................ 25
Figure 16. RSD at the end of HVS trafficking.............................................................................. 26
Figure 17. In-depth elastic deflections under 40-kN test load, MDD 7........................................ 26
Figure 18. In-depth elastic deflections under 80-kN test load, MDD 7........................................ 27
Figure 19. In-depth elastic deflections under 100-kN test load, MDD 7...................................... 27
Figure 20. Contribution of pavement layers to surface elastic deflection based on 40-kN load,
MDD 7. ................................................................................................................................. 29
Figure 21. Profile of elastic deflections with depth. ..................................................................... 29
v
Figure 22. Elastic deflections versus test load level at pavement depth of 250 mm, MDD 7. ..... 30
Figure 23. Summary of back-calculated moduli from in-depth deflections. ................................ 32
Figure 24. In-depth elastic deflections for Section 502 (tested under dry conditions). ................ 32
Figure 25. Surface crack schematics for Section 543. ................................................................. 34
Figure 26. Density of cracking on Section 543 after completion of HVS trafficking. ................. 36
Figure 27. Progression of crack density on Section 543............................................................... 37
Figure 28. Summary of FWD elastic deflections.......................................................................... 38
Figure 29. Results of back-calculation, asphalt concrete layer with and without ATPB layer. .. 40
Figure 30. Results of back-calculation, aggregate base layer. ..................................................... 40
Figure 31. Results of back-calculation, ATPB layer. .................................................................. 42
Figure 32. Results of back-calculation, subgrade. ....................................................................... 42
Figure 33. NEED .......................................................................................................................... 44
Figure 34. Variation in bound pavement layers as measured with Ground Penetrating Radar
(GPR). ................................................................................................................................... 44
Figure 35. Reduction in bound layer thickness across Section 543 as measured with GPR. ....... 45
Figure 36. Locations of forensic activities on Section 543........................................................... 46
Figure 37. Core extracted from centerline at Station 13, Section 543.......................................... 47
Figure 38. Core extracted from outside the trafficked area at Station 13, Section 543. .............. 48
Figure 39. Core extracted from centerline at Station 9, Section 543............................................ 50
Figure 40. ATPB with and without fines infiltration................................................................... 51
Figure 41. Sites of field percolation tests...................................................................................... 53
Figure 42. Percolation tests on aggregate base. ............................................................................ 55
Figure 43. Percolation tests on ATPB........................................................................................... 55
vi
Figure 44. Average air-void contents across Section 543 after HVS trafficking. ........................ 56
Figure 45. Dynamic Cone Penetrometer (DCP) tests at Station 5, Section 543........................... 58
Figure 46. Dynamic Cone Penetrometer (DCP) tests at Station 13, Section 543......................... 58
Figure 47. Dynamic Cone Penetrometer (DCP) tests in trench, Section 543. .............................. 59
Figure 48. Layer profiles measured in trench, south face of trench at Station 7, Section 543. .... 61
Figure 49. Layer profiles measured in trench, north face of trench at Station 9, Section 543. .... 62
Figure 50. Summary of layer thicknesses across transverse distance of trench. .......................... 63
Figure 51. Methodology followed in the fatigue analysis system to determine ESALs............... 68
vii
viii
LIST OF TABLES
Table 1 Pavement Layer Thicknesses for Section 543 ............................................................ 5
Table 2 HVS Loading Sequence for Section 543 .................................................................. 10
Table 3 Data Collection Program........................................................................................... 11
Table 4 Air and Pavement Temperatures for Section 543 ..................................................... 14
Table 5 Depths of Moisture Content Measurements ............................................................. 17
Table 6 Average Rate of Rutting at Three Traffic Loads ...................................................... 19
Table 7 HVS Traffic for Previous Sections Tested under Dry Conditions............................ 23
Table 8 Percent Contribution of Pavement Layers to Surface Deflection............................. 28
Table 9 D1 Normalized Deflections for Section 543............................................................. 38
Table 10 Summary of Layers and Thickneses Considered for Back-calculation .................... 39
Table 11 Description of Test Sections Used for Percolation Tests.......................................... 52
Table 12 Summary of Air-Void Contents for Section 543 ...................................................... 57
Table 13 Summary of DCP Penetration Rates, Section 543.................................................... 59
Table 13 Summary of Pavement Structures for Analysis ........................................................ 66
Table 14 Summary of Pavement Responses under 40-kN Load ............................................. 66
Table 15 Summary of Calculation of ESALs Using UCB Fatigue Analysis System.............. 67
ix
x
EXECUTIVE SUMMARY
This report is the first in a series of four reports that describe the results of accelerated
pavement tests on full-scale pavements with “wet” base conditions at the Pavement Research
Center, located at the University of California Berkeley Richmond Field Station (RFS). The
report contains a summary of the results and associated analysis of a pavement section comprised
of three lifts of asphalt concrete, an asphalt treated permeable base (ATPB) layer, and untreated
aggregate base layers on top of a prepared subgrade and designated as Section 543. The
pavement section is a termed a “drained’ pavement due to the inclusion of an ATPB layer
between the asphalt concrete layers and the untreated aggregate base. Tests on this test section
have been performed as part of the Goal 5 Accelerated Test Program for the evaluation of
drained and undrained pavements under conditions of water infiltration.
The main objectives of the test program are:
• Develop data to quantitatively compare and evaluate the performance of drained and
undrained asphalt concrete pavements under “wet” conditions. Wet conditions intend
to simulate approximate surface water infiltration rates that would occur in a badly
cracked asphalt concrete layer along the north coast of California during a wet month.
• Measure the effectiveness of the drained pavement in mitigating surface water
infiltration and thereby preventing a decrease in stiffness and strength of the unbound
layers.
Other objectives include:
• Quantify elastic moduli of the pavement layers,
• Quantify the stress dependence of the pavement layers, and
• Determine the mechanics of failure of the pavement structure.
xi
Heavy Vehicle Simulator (HVS) testing was begun in May 1999 and was completed in
March 2000 after the application of more than 1.1 × 106 load repetitions. At the end of the test,
the pavement section had more than 20 mm of surface rutting, and a surface crack density of 2.5
m of cracks per square meter (m2) of pavement.
Chapter 2 describes the test program for Section 543. Design and as-constructed
thicknesses of the pavement components were as follows:
Layer Design Thickness As-Constructed Thickness Asphalt Rubber Hot Mix 40 mm 37 mm Dense Graded Asphalt Concrete 148 mm 141 mm Asphalt Treated Permeable Base 75 mm 69 mm Aggregate Base 182 mm 181 mm Aggregate Subbase 215 mm 187 mm
The test program was conducted in four stages: Stage 1, definition of the initial condition
of the section under conditions of no water infiltration; Stage 2, establishing the initial structural
condition of the section under water infiltration; Stage 3, HVS trafficking under conditions of
water infiltration; and Stage 4, definition of the structural condition of the section after HVS
trafficking.
Data collection for Stages 1, 2, and 4 were based on the deflection measurements
obtained with a Dynatest Falling Weight Deflectometer (FWD), and moisture content
measurements. Stage 4 included some destructive tests to verify the results obtained during Stage
3.
Data collection for Stage 3 is summarized in Table 3. Loading, applied by dual radial
tires inflated to a pressure of 720 kPa, consisted of 236,000 repetitions of a 40-kN load followed
by 332,000 repetitions of an 80-kN load, followed by 629,000 repetitions of a 100-kN load. At
xii
the termination of loading, surface rutting and fatigue cracking were visible throughout the test
section. Lateral wander of the wheels was distributed across the 1-m width of the test section.
Pavement responses were obtained using Multiple Depth Deflectometers (MDD), the
Road Surface Deflectometer (RSD), and the laser profilometer. Fatigue cracking was
photographically recorded and analyzed using a digital image analysis procedure.
Thermocouples were used to measure air and pavement temperatures at various depths in the
asphalt concrete. To maintain a consistent temperature around 20ºC, a temperature control
cabinet was installed. Chapter 3 summarizes the data obtained during the course of loading and
associated analyses.
Section 543 rapidly failed due to surface rutting during trafficking at the 100-kN load.
Failure resulted form stripping in the ATPB layer and loss of adhesion in the prime coat placed
on top of the aggregate base. Stripping was caused by the high shear stresses experienced under
the 100-kN load present in the ATPB layer and water infiltration, which softened the binder in
the ATPB. As the binder gradually washed out of the ATPB, the ATPB started losing its stiffness
and at the same time fine particles from the aggregate base infiltrated the voids in the ATPB.
Results from back-calculation of FWD tests indicated that the stiffness of the ATPB was reduced
to that of an untreated aggregate base. Percolation tests indicated that the permeability of the
ATPB in the trafficked area was reduced from 1 cm/s to 0.001 cm/s. This prevented water from
draining from the ATPB, thus contributing to the stiffness reduction.
Results of the analysis reported in Chapter 4 indicate that if integrity of the ATPB had
been maintained, the section would likely have failed by fatigue cracking rather than surface
rutting after about 2,300 million ESALs at a reliability level of 50 percent. Actually, only 35
million ESALs were applied to the test section based on the use of the Caltrans load equivalency
xiii
factor of 4. An analysis was also conducted on a pavement section in which the ATPB layer was
replaced by an equivalent thickness of DGAC; i.e., 40 mm of DGAC was substituted for the 75
mm of ATPB. Analysis of this pavement section using the same procedure as above indicated
that this equivalent section would sustain about 3,300 million ESALs prior to failure in fatigue,
also at a 50 percent reliability level. Internal drainage was not included in the alternate structure
to intercept water entering through the pavement surface. Entry of water from the surface can be
mitigated, as noted in earlier studies, by properly compacting the asphalt concrete and using
sufficient thickness to mitigate the potential for load associated cracking. In addition, by
increasing the degree of compaction in the aggregate base and subbase, the contribution of
rutting from these layers would be minimal, all of which would eliminate the need for the ATPB.
The analysis indicated that surface rutting in Section 543 could be attributed to the
reduced stiffness of the ATPB layer. As a result, rutting occurred both in the ATPB and the
aggregate base. The stiffnesses of the untreated aggregate base and subbase are dependent on the
stress state. Reduced stiffness of the ATPB resulted in an increase in the stress in the aggregate
base, thus contributing to its increased rutting propensity and likely contributing to the observed
surface rutting as well.
If ATPB is to remain effective as a drainage layer, it will be necessary to ensure that the
material does not strip and that an adequate filter be provided adjacent to the ATPB to minimize
the intrusion of fines. In addition, edge and transverse drains need to be maintained in order to
prevent clogging. If not, water will be retained in the ATPB resulting in the type of performance
observed in this test.
Chapter 5 contains conclusions, which are briefly summarized as follows:
xiv
The ATPB failed due to stripping and the prime coat did not serve as an adequate barrier to
prevent fines from the aggregate base from intruding into the ATPB. Stripping was caused by the
high shear stress produced by the 100-kN load and presence of water. The ATPB layer became
clogged with fines from the underlying aggregate base, thereby reducing the permeability of the
ATPB layer from 1 cm/s to 0.001 cm/s. This created an increased degree of saturation within the
pavement that quickly contributed to deterioration in the performance of the pavement section.
Section 543 failed initially due to surface rutting. The reduced stiffnesses in the ATPB
and aggregate base contributed to the rapid increase in surface rutting. The reduced stiffnesses of
these layers also resulted in increased tensile strain in the asphalt concrete layers. This in turn
resulted in a rapid increase in surface cracking as well.
xv
xvi
1.0 INTRODUCTION
This report presents the data from the accelerated pavement testing (APT) program
conducted with the Heavy Vehicle Simulator (HVS) on test Section 543. This section is part of
the APT program described in the test plan for CAL/APT Goal 5, “Performance of Drained and
Undrained Flexible Pavement Structures under Wet Conditions.”(1) This document provides
detailed information on the Goal 5 program.
Three sections were available from the Goal 3 study for HVS loading as part of the Goal
5 investigation. Section 543 had a 38-mm ARHM-GG (asphalt rubber hot mix—gap graded)
surface course and an ATPB (asphalt treated permeable base) drainage layer. Section 544 had a
38-mm ARHM-GG surface course and only the untreated aggregate base layer (undrained).
Section 545 had a 75-mm DGAC (dense graded asphalt concrete) surface course and only the
untreated aggregate base layer (undrained).
The objective for the test on Section 543 was to evaluate the performance of a typical
“drained” pavement section under wet conditions. The “drained” pavement section in this case is
comprised of a 75-mm layer of ATPB between the asphalt concrete and the aggregate base and a
perforated pipe drainage system at the shoulder. An “undrained” pavement is a conventional
pavement with a Class 2 untreated aggregate base and subbase, and without an ATPB layer.
The purpose of the ATPB layer and drainage system is to intercept water entering the
pavement either through cracks in the asphalt concrete or through very permeable asphalt
concrete and carry it out of the pavement before it reaches the unbound layers, where it can
significantly reduce their strength and stiffness characteristics.
Wet conditions for Section 543 were intended to simulate approximate surface infiltration
rates that would occur along the north coast climate region in California (2) during a wet month
for a badly cracked asphalt concrete layer. Because the surface course of Section 543 was
1
initially uncracked, water was introduced through small holes drilled through the asphalt
concrete and into the ATPB. This was intended to simulate water infiltrating from adjacent lanes
or from water entering pavement constructed in cuts.
1.1 Objectives
The main objective of this test program was to evaluate the effectiveness of drained
pavements in preventing a decrease in stiffness and strength of unbound materials (base,
subbase, and subgrade) due to an increase in water content produced by water infiltration. Other
objectives included:
• Comparison of the long term performance of drained and undrained pavements under
wet conditions;
• Quantification of elastic moduli of the various pavement layers using deflections
obtained from a slow moving wheel and from the falling weight deflectometer;
• Determination of the failure mechanism(s) of the pavement sections; and
• Evaluation of the effectiveness of non-destructive and partially destructive methods
for assessing the pavement structural condition.
An earlier report of a laboratory study together with individual reports covering the
results of HVS loading on each of the three test sections and a summary and evaluation of the
results of the three sections will complete the Goal 5 study.(3–5)
1.2 Test Sequence
The sequence of HVS testing of the three test sections was 1) Section 543 (drained), 2)
Section 544 (undrained), and 3) Section 545 (undrained). This report presents the results of the
loading tests on Section 543.
2
1.3 Organization of Report
Chapter 2 of the report contains a description of the test program for Section 543,
including the loading sequence, instrumentation, and data collection scheme. Chapter 3 presents
a summary and analysis of the data collected during the test, including back-calculated moduli
from deflection tests and their effectiveness to establish structural condition. Chapter 4 presents
the mechanistic pavement analyses used to evaluate the comparative performance of the three
test sections. Chapter 5 contains a summary of the results and conclusions.
3
4
2.0 TEST PROGRAM
The following sections present the test section layout, environmental conditions
encountered during the test, instrumentation used to collect data, and the test program.
2.1 Test Section Layout
Section 543 is 8 m long by 1 m wide, divided into sixteen 0.5-m long “stations,” as
shown in Figure 1. The pavement was constructed with a two-percent cross slop and 0.5-percent
longitudinal slope in all layers above the aggregate subbase. The design thickness, pavement
layers, and as-built layer thicknesses are summarized in Table 1. The as-built pavement
thicknesses were obtained from direct measurements on the non-trafficked portion of a trench
dug in the section at the conclusion of HVS loading.
Table 1 Pavement Layer Thicknesses for Section 543
Layer Design Thickness (mm)
As-Constructed Thickness (mm)
Asphalt Rubber Hot Mix 40 37 Dense Graded Asphalt Concrete 148 141 Asphalt Treated Permeable Base 75 69 Aggregate Base 182 181 Aggregate Subbase 215 187
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
X X X X X
X
Stations
MDD Location
RSD Measurement Location
Water Infiltration Hole
Turnaround Zone
Traffic Zone
1 m
Cross Slope
2 % 0.5 %
0.75
m
OC
0.5 m OC
8 m
Figure 1. Section 543 layout and locations of MDD and RSD measurements.
5
2.2 Environmental Conditions
The accelerated pavement test was conducted under wet conditions at a pavement of
about 20ºC. Water infiltration into the treated permeable base was controlled at a rate of 9.7 liters
per hour over an area 7.4 m wide by 12 m long, determined to be representative of the rate of
water infiltration into a pavement during an average peak week in Eureka, California of 51.3 mm
rainfall.(2)
The water was introduced through a line of holes parallel to and upslope from the HVS
test section, as shown in Figures 1 and 2. An electronic valve and timer system continuously
controlled water ingress to the pavement during the test.
For the test, the actual target pavement temperature at a pavement depth of 50 mm was
20 ± 2ºC. This was maintained by a temperature control box surrounding the HVS.
2.3 Instrumentation
Instrumentation and measuring equipment used on Section 543 included:
• Multiple Depth Deflectometers (MDDs) to measure deflection and permanent
deformation at various depths in the pavement section;
• a Road Surface Deflectometer (RSD) to measure surface deflections,
• a laser profilometer to monitor surface rutting;
• thermocouples to monitor pavement temperatures; and
• hydroprobes and tensiometers to monitor moisture content in the unbound layers.
Figure 1 shows the locations of the MDDs in the test section as well as the locations for
the RSD measurements. Laser profilometer measurements were obtained at each of the
numbered stations. Figure 2 contains a cross section of the pavement illustrating the locations of
6
36 mm
0 mm
643 mm
176 mm
240 mm
420 mm
Asphalt Rubber
Dense GradedAsphalt Concrete
Asphalt TreatedPermeable Base
Aggregate Base
Aggregate Subbase
Subgrade
MDD Thermocouples
0 mm
38 mm
104 mm
170 mm
245 mm
0 mm
250 mm
430 mm
645 mm
850 mm
Waterinfiltration
holes
Figure 2. Depths of MDDs, thermocouples, and water infiltration holes in Section 543.
7
the MDDs, thermocouples, and the depth of the water infiltration holes. For a complete
description of the instrumentation used for this study, refer to Reference (6).
2.4 Test Program
The test program consisted of four stages, as shown in Figure 3. These stages can be
summarized as follows:
• Stage 1: Prior to introduction of water, water content determinations and FWD tests
were conducted to establish initial moisture conditions and structural response
• Stage 2: During the water infiltration period, water content measurements and FWD
tests were conducted to establish moisture conditions and the structural response
before HVS testing.
• Stage 3: The pavement was subjected to HVS loading with water infiltrating the
pavement. Pavement monitoring continued throughout the loading period using the
instrumentation and test equipment described in Section 2.3.
• Stage 4: After HVS trafficking, a detailed forensic study was conducted which
included FWD tests, trenching, and sampling of the pavement materials for laboratory
tests.
2.4.1 Falling Weight Deflectometer Testing
FWD tests were conducted to monitor surface deflections and changes in the layer
moduli of the pavement section by back-calculation procedures. Loads of 20, 40, and 60 kN were
used with the deflection sensors (geophones) spaced at 0, 200, 300, 600, 900, 1200, and 1500
mm from the loading plate. Surface deflections for the various loads were normalized to a 40-kN
load and back-calculated moduli were determined for this load condition.
8
ID Task Name1 Stage 1: No water infiltration
2 Section 543
3 Stage 2: Water infiltration
4 Section 543
5 Stage 3: HVS Testing and water infiltration
6 Section 543
7 Stage 4: Forensic Activities
8 Section 543
Apr May
Section
Figure 3. Schedule for test program.
9
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May1999
543 Test Program Schedule
2.4.2 Heavy Vehicle Simulator Trafficking
For trafficking of Section 543, the HVS was equipped with dual truck tires, representing
one half of a single axle. Load was applied through two Goodyear radial tires G159 11R 22.5)
inflated to a pressure of 720 kPa. Table 2 presents the trafficking loads, loading sequence, and
the associated HVS load repetitions.
Table 2 HVS Loading Sequence for Section 543 HVS Load (kN) HVS Repetitions 40 0 to 236,588 80 236,588 to 568,356 100 568,356 to 1,197,685
The test wheel trafficked the entire length of the 8-m long test section. Lateral wander
over the 1-m width of the test section was programmed to simulate representative traffic wander
in a typical highway lane. The 1-m sections at both ends of the trafficked area (Stations 0–2 and
14–16 in Figure 1) serve as “turnaround zones” in which the test wheel decelerates and
accelerates. Pavement performance was evaluated for the 6 m length between these turnaround
zones (Station 2–14) in which the HVS wheel speed is constant.
2.4.3 Data Collection
Table 3 summarizes the data collection sequence for Section 543.
10
Table 3 Data Collection Program HVS
Repetitions Laser
Profilometer Multi-Depth
Deflectometer Road Surface Deflectometer
Crack Monitoring
0 X 40 kN 40 kN 12005 X 40 kN 40 kN 28881 X 40 kN 40 kN 46323 X 40 kN 40 kN 58791 X 40 kN 40 kN 89619 X 40 kN 40 kN 124341 X 40 kN 40 kN 159071 X 40 kN 40 kN 197535 X 40 kN 40 kN 236588 X 40 kN 40 kN 285201 X 40 kN 40 kN 331855 X 40 kN 40 kN 370225 X 40 kN 40 kN 406098 X 40 kN 40 kN 457197 X 40 kN 40 kN 496169 X 40 kN 40 kN 568356 X 40, 80, 100 kN 40, 80, 100 kN 638196 X 40, 100 kN 40, 100 kN 694154 X 40, 100 kN 40, 100 kN 779237 X 40, 100 kN 40, 100 kN 856127 X 40, 100 kN 40, 100 kN 888069 927069 X 40, 100 kN 40, 100 kN 939343 40, 100 kN X 994171 X 40, 100 kN 40, 100 kN 1057304 X 40, 100 kN 40, 100 kN X 1114295 X 40, 100 kN 40, 100 kN X 1197685 X 40, 100 kN 40, 100 kN X
11
12
3.0 SUMMARY OF TEST DATA
This chapter provides a summary of the test data collected for during the four stages of
testing. Test data include:
• Stage 1: temperature, moisture content, and FWD data
• Stage 2: temperature, moisture content, and FWD data
• Stage 3: temperature, moisture content, permanent deformation, elastic deflection,
and crack progression data
• Stage 4: FWD and forensic study data.
3.1 Temperature and Moisture Condition Data
Temperature and water content data were gathered throughout the test program for
Section 543. These data are summarized in the following sections.
3.1.1 Temperature Data
This section includes a summary of air and pavement temperatures during FWD tests and
HVS loading.
Air temperatures were automatically collected during FWD tests by sensors incorporated
in the FWD equipment. Pavement temperatures were estimated from air temperatures using
Bell’s equation. Average air and pavement temperatures during FWD tests conducted during
Stages 1, 2, and 4 are summarized in Table 4. Additional FWD data with temperature are
presented in Section 3.6.
13
Table 4 Air and Pavement Temperatures for Section 543 Average Temperature (°C) Stage of
Testing Equipment Air Pavement 1 FWD on-board sensor 24 19 2 FWD on-board sensor 21 17 3 HVS thermocouples 18 20 4 FWD on-board sensor 15 14
Figure 4 shows average daily temperatures inside Building 280 at the University of
California Berkeley Richmond Field Station, where the HVS was located, and average air
temperatures inside the HVS temperature control box. The average daily temperatures were
calculated from hourly temperatures recorded during testing. Average air temperatures
monitored inside the temperature control box were uniform with minimal variations. The average
air temperature inside the temperature control box during testing was 18.3ºC.
Average daily asphalt concrete temperatures for Section 543 are shown in Figure 5.
These temperatures were recorded at various depths in the pavement: overlay, upper lift of the
asphalt concrete, lower lift of the asphalt concrete, and ATPB layer. Generally, the asphalt
concrete temperatures were fairly uniform and varied only slightly with air temperature inside
the temperature control box. Average temperatures in the asphalt concrete layers were
approximately 6 percent higher than the average air temperatures inside the control box with an
overall average of about 19.5ºC.
3.1.2 Rainfall and Moisture Content Data
Figure 6 shows average monthly rainfall data collected at the National Weather Service
weather station in Richmond, California. Figures 7 and 8 show average volumetric moisture
contents for the aggregate base, subbase, and subgrade materials obtained using a hydro-probe
test device. Depths of moisture content measurements are summarized in Table 5.
14
0
5
10
15
20
25
16-N
ov-9
9
23-N
ov-9
9
30-N
ov-9
9
07-D
ec-9
9
14-D
ec-9
9
21-D
ec-9
9
28-D
ec-9
9
04-J
an-0
0
11-J
an-0
0
18-J
an-0
0
25-J
an-0
0
01-F
eb-0
0
08-F
eb-0
0
15-F
eb-0
0
22-F
eb-0
0
29-F
eb-0
0
Date
Tem
pera
ture
, ºC
Inside Temperature Control Box
Outside Temperature Control Box
Figure 4. Average air temperatures inside and outside HVS temperature control box, Section 543.
15
17
19
21
23
25
16-N
ov-9
9
23-N
ov-9
9
30-N
ov-9
9
07-D
ec-9
9
14-D
ec-9
9
21-D
ec-9
9
28-D
ec-9
9
04-J
an-0
0
11-J
an-0
0
18-J
an-0
0
25-J
an-0
0
01-F
eb-0
0
08-F
eb-0
0
15-F
eb-0
0
22-F
eb-0
0
29-F
eb-0
0
Date
Tem
pera
ture
, ºC
Surface38 mm104 mm170 mm245 mm
Figure 5. Pavement temperatures at depth, Section 543.
15
0
2
4
6
8
10
12
14
16
18
20
Apr
-99
May
-99
Jun-
99
Jul-9
9
Aug
-99
Sep
-99
Oct
-99
Nov
-99
Dec
-99
Jan-
00
Feb-
00
Mar
-00
Apr
-00
May
-00
Jun-
00
Jul-0
0
Date
Prec
ipita
tion,
cm
N/A
*
Stage 1Stage 1 Stage 2Stage 2 Stage 3Stage 3 Stage 4Stage 4
*Note: Data was not available for Richmond, California weather station for the month of February 2000. Data shown here for February 2000 is from nearby National Climate Data Center weather station in Berkeley, California. Figure 6. Average monthly rainfall during testing of Section 543.
0
1
2
3
4
5
6
7
06/0
9/99
06/2
9/99
07/1
9/99
08/0
8/99
08/2
8/99
09/1
7/99
10/0
7/99
10/2
7/99
11/1
6/99
12/0
6/99
12/2
6/99
01/1
5/00
02/0
4/00
02/2
4/00
03/1
5/00
04/0
4/00
Date
Moi
stur
e C
onte
nt, P
erce
nt
Aggregate Base
Aggregate Subbase
Stage 1 Stage 2 Stage 3
Figure 7. Volumetric moisture content in aggregate base and subbase layers.
16
25
30
35
06/0
9/99
06/2
9/99
07/1
9/99
08/0
8/99
08/2
8/99
09/1
7/99
10/0
7/99
10/2
7/99
11/1
6/99
12/0
6/99
12/2
6/99
01/1
5/00
02/0
4/00
02/2
4/00
03/1
5/00
04/0
4/00
Date
Moi
stur
e C
onte
nt, P
erce
nt
940 mm1240 mm1320 mmSeries1
Stage 1 Stage 2 Stage 3
Figure 8. Volumetric moisture content in the subgrade at various depths from the surface.
Table 5 Depths of Moisture Content Measurements Unbound Layer Depth from Surface, mm Aggregate Base 468 and 508 Aggregate Subbase 667 and 707 Subgrade 940, 1240, and 1320
Moisture contents of the base and subbase, Figure 7, were averaged across measurements
taken at those layers. These data show that the moisture content of the aggregate base and
subbase were not affected by the water that infiltrated from the surface through the infiltration
holes. It is possible that the permeability of the aggregate base was sufficiently small to prevent
water intrusion. Moisture content data for the subgrade, Figure 8, show variations that followed
the precipitation trends. The subgrade soil at the test site was not isolated from the surrounding
soil outside of Building 280, thus, being subjected to the same seasonal variations.
17
Figure 9 shows moisture contents from aggregate samples extracted from two boreholes
drilled at the end of the test program. The two boreholes were located at Station 5 and 13.1
Volumetric moisture contents were estimated from gravimetric water contents using a
bulk density in the aggregate base of 2.200 kg/m3. This bulk density was obtained using the
sand-cone method. These data indicate higher volumetric moisture contents for the sampled
material than those obtained with the hydroprobe.
3.2 Permanent Deformation
Permanent deformation data were gathered using the laser profilometer and multi-depth
deflectometers (MDDs). The following sections summarize these data.
200
300
400
500
600
4 5 6 7 8 9 10 11 12Gravimetric Water Content, Percent
Dep
th fr
om S
urfa
ce, m
m
200
250
300
350
400
450
500
550
6008.8 10.8 12.8 14.8 16.8 18.8 20.8 22.8 24.8
Volumetric Water Content, Percent
Borehole @ (-500mm, Station 5)
Borehole @ (-500mm, Station 13)
Figure 9. Moisture contents from aggregate samples extracted during forensic study.
1 Figure 3.6 in Section 3.8, which discusses the forensic studies, shows locations of all of the boreholes placed in the test section. These two boreholes were located at a distance of ~500 mm, as shown in Figure 3.6.
18
3.2.1 Surface Rutting Measured with the Laser Profilometer
Figure 10 shows accumulated average maximum rut depth for Section 543. A rapid rate
of permanent deformation is note under the 100-kN trafficking load compared to the 40-kN and
80-kN trafficking loads. Table 6 summarizes average rates of rutting for each of the three levels
of traffic.
Figure 11 shows the rut depth distribution at the end of the 40-, 80-, and 100-kN
trafficking loads. The rut distribution is fairly uniform throughout the section with greater rutting
localized around Stations 3 and 13. The center of the HVS wheelpath is marked as 0 mm
transverse distance.
Table 6 Average Rate of Rutting at Three Traffic Loads
Trafficking Load
Rate of Rutting (mm/million load repetitions)
40 kN 3.8 80 kN 7.7 100 kN 30.0
0
5
10
15
20
25
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000
HVS Load Applications
Rut
Dep
th, m
m
Section 543 Drained
Average Maximum Rut Depth Failure Criterion = 12.5 mm
40 kN Load 80 kN Load 100 kN Load
Figure 10. Average maximum rut depth for Section 543.
19
0 1000 2000 3000 4000 5000 6000 7000 8000-1000
-500
0
500
1000
Distance (mm)
Dis
tanc
e (m
m)
Section 543, 236k repetitions
-2
0
00
0
000
0 0
0 0
00000
000
0
0
00
0
00
0
0
00
00
0
00
00
0
0
0
2468
0 1000 2000 3000 4000 5000 6000 7000 8000-1000
-500
0
500
1000
Distance (mm)
Dis
tanc
e(m
m)
Section 543, 568k repetitions
-5-4
-4-4
-3-3
-3
-3
-3
-3
-3
-3
-2
-2
-2-2
-2-2 -2 -2
-1 -1-1
-1-1 -1 -1
-1-1-1-1
-1
-1
000
000
0
0
0
000
000 0
000
0
0 0
0
0
00
0
0
0
0
00
0
0
1
1
1
0 1000 2000 3000 4000 5000 6000 7000 8000-1000
-500
0
500
1000
Distance (mm)
Dis
tanc
e (m
m)
Section 543, 1197k repetitions
-25
-20 -20 -20 -20
-20-20-20
-15
-15 -15 -15
-15-15-15
-10 -10 -10 -10
-10-10-10
-10
-5
-5 -5 -5
-5
-5-5-5
000 00
0
00 0 0 0 0
00
0 0 00 0 00
00
0
0 0
0
0
Figure 11. Section 543 surface rut profiles at various stages of HVS trafficking.
20
3.2.2 In-depth Permanent Deformation
Figure 12 shows the permanent deformation recorded at several layers in the pavement
using the MDD (refer to Figure 1 for depths of MDDs in the pavement). The deformation data at
the top of the subgrade are not show because the MDD installed at this location was not
functioning during HVS trafficking.
The rapid increase in permanent deformation under the 100-kN load with load
applications is evident for all the layers. The MDD sensor positioned at the ATPB and AB
interface (250 mm depth) recorded the highest increase in permanent deformation among all
layer interfaces. Cores obtained along the centerline of the test section indicate that the increase
in permanent deformation under conditions of water infiltration and high loads at the ATPB and
AB interface can be attributed to the following:
• stripping of the ATPB layer,
• intrusion of fines from the AB into the ATPB, and
• softening of the AB layer.
Figure 13 shows the permanent deformation attributed to each pavement layer with
increase in HVS load applications. It can be seen from the figure that most of the rutting
occurred in the upper layers (asphalt layers and granular base). The asphalt concrete layers and
the ATPB significantly contributed to rutting during the trafficking of the 40-kN load. However,
as the trafficking load increased, the contribution of the unbound granular material to the total rut
depth increased significantly.
3.3 Comparisons of the Performance of Similar Test Sections with Dry Base Conditions
To provide some comparisons between the behavior of Section 543 and comparable
sections tested earlier under dry conditions, the test results from Section 543 will be compared to
21
0
5
10
15
20
25
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000
HVS Load Applications
Def
orm
atio
n, m
m
0 mm, Surface
250 mm, Aggregate Base
430 mm, Aggregate Subbase
850 mm, Subgrade
40 kN 80 kN 100 kN
Figure 12. In-depth permanent deformation in Section 543.
0
10
20
30
40
50
60
70
80
90
100
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000
HVS Load Applications
Con
trib
utio
n to
Sur
face
Rut
ting,
per
cent Asphalt bound + ATPB
Aggregate Base
Below Aggregate
40 kN 80 kN 100 kN
Figure 13. Contribution of pavement layers to surface rutting in Section 543.
22
the test results from the initial Goal 1 Test Sections 500 and 502 and Goal 3 Test Sections 514
and 515. The two Goal 1 sections were the original sections containing ATPB and tested in the
dry condition. There was some difference in response of these two sections due to changes in
subgrade water content. Sections 514 and 515 corresponded to Sections 500 and 502, which had
been overlaid with layers of DGAC and ARHM-GG, respectively. Summaries of HVS loadings
for the Goal 1 and Goal 3 test sections are included in Table 7.
Table 7 HVS Traffic for Previous Sections Tested under Dry Conditions Goal 1 Load Applications (× 103) Goal 3 Load Applications (× 103) Section
(Goal 1/Goal 3) 40 kN 80 kN 100 kN 40 kN 80 kN 100 kN 500/514 150 50 2370 172 145 1350 502/515 150 50 2470 128 218 2065
Figure 14 shows in-depth permanent deformation data obtained with the MDDs for the
Goal 1/Goal 3 HVS test sections. Locations of the MDDs are similar to those for Section 543,
i.e., close to the surface, top of the aggregate base, top of the aggregate subbase, and top of the
subgrade.
A comparison of Figure 12 with Figure 14 indicates a significant difference in rutting
performance of the layers. Less surface rutting was observed in Sections 500 and 502 even
though more than 4 million HVS load applications were applied. Is addition, the sections barely
approach the surface rutting failure criterion of 12.5 mm after the two HVS programs. Rutting in
the aggregate base layers and in the subgrade were less than 5 mm and 2 mm respectively after
both HVS programs. In contrast, Section 543 reached the surface rutting failure criterion of 12.5
mm after about 850 thousand HVS load applications. At this level of rutting, about 8 mm of
rutting occurred on top of the aggregate base. These results point to a significant difference in
performance of the drained section in the dry state versus that in the wet condition.
23
0
5
10
15
20
25
- 1,000,000 2,000,000 3,000,000 4,000,000 5,000,000 6,000,000
Load Applications
Def
orm
atio
n, m
m500/514 Surface502/515 Surface500/514 Top Aggregate Base502/515 Top Aggregate Base500/514 Top Aggregate Subbase502/515 Top Aggregate Subbase500/514 Top Subgrade502/515 Top Subgrade
Goal 1 HVS Program Goal 3 HVS Program
Figure 14. In-depth permanent deformation from Goal 1/Goal 3 test sections.
3.4 Elastic Deflections
Elastic deflections in the pavement sections were measured with the Road Surface
Deflectometer (RSD) and MDDs. Surface and in-depth deflection data presented in the following
sections represent peak values obtained from the deflection response of the pavement.
3.4.1 Surface Elastic Deflection Data
Surface elastic deflections were monitored using the RSD along the section centerline at
Stations 4, 6, 8, 10, and 12 under the 40-, 80-, and 100-kN test loads. Figure 15 shows average
deflection data for the three test loads. Elastics deflections progressively increased under the 80-
kN and 100-kN trafficking loads, with the highest rate of increase under the 100-kN trafficking
24
0
500
1000
1500
2000
2500
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000HVS Load Applications
Def
lect
ion
(mm
×10
-6)
40 kN80 kN100 kN
Test Load
40 kN Load 80 kN Load 100 kN Load
Figure 15. RSD deflections for Section 543.
load. No significant increase in elastic deflections was recorded during trafficking under the 40-
kN load.
Figure 16 shows RSD deflections along Section 543 at the completion of HVS testing.
The highest deflections were measured around Station6, while the lowest deflections were
measured near Station 12.
3.4.2 In-depth Elastic Deflections
In-depth elastic deflection data under the 40-, 80-, and 100-kN test loads are presented in
Figures 17–19. Response of the pavement layers under the 40-kN test load, Figure 17, remained
relatively constant during trafficking at 40 kN. Under the trafficking at 80 and 100 kN,
progressive increases in elastic deflections were recorded in all of the layers with the highest rate
25
0
500
1000
1500
2000
2500
0 2 4 6 8Location Along Test Section, m
Elas
tic S
urfa
ce D
efle
ctio
n, m
m ×
10-6
40 kN80 kN100 kN
Test Load
Figure 16. RSD at the end of HVS trafficking.
0
500
1000
1500
2000
2500
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000HVS Load Applications
Def
lect
ion,
m ×
10-6
0 mm, Surface250 mm, Aggregate Base430 mm, Aggregate Subbase850 mm, Subgrade
MDD Depth
40 kN Load 80 kN Load 100 kN Load
Figure 17. In-depth elastic deflections under 40-kN test load, MDD 7.
26
0
500
1000
1500
2000
2500
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000
HVS Load Applications
Def
lect
ion,
m ×
10-6
0 mm, Surface250 mm, Aggregate Base430 mm, Aggregate Subbase850 mm, Subgrade
MDD Depth
40 kN Load 80 kN Load 100 kN Load
Figure 18. In-depth elastic deflections under 80-kN test load, MDD 7.
0
500
1000
1500
2000
2500
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000HVS Load Applications
Def
lect
ion,
m ×
10-6 0 mm, Surface
250 mm, Aggregate Base430 mm, Aggregate Subbase850 mm, Subgrade
MDD Depth
40 kN Load 80 kN Load 100 kN Load
Figure 19. In-depth elastic deflections under 100-kN test load, MDD 7.
27
of increase observed with the 100-kN load. The data also show that the rate of increase of the
elastic deflections leveled off at approximately 950,000 load applications when surface cracking
appeared.
The percent contribution of each pavement layer to elastic surface deflection is presented
in Figure 20 for the 40-kN test load. Table 8 summarizes the percent contribution of each
pavement layer to surface deflections under each trafficking level under all test load levels. On
average, 80 percent of the total surface deflection can be attributed to the unbound layers.
Table 8 Percent Contribution of Pavement Layers to Surface Deflection Average Percent Contribution to Surface Deflection at Trafficking Load
Pavement Layer
40 kN 80 kN 100 kN Asphalt Layers 16 24 20 Aggregate Base 29 22 25 Below Aggregate Base 56 53 55
Figure 21 presents the deflection data as a profile of deflections with respect to pavement
depth. The reduction in slope of the upper layers (asphalt bound and aggregate base) indicates a
reduction in stiffness of these layers. Based on the performance data, the reduction in stiffness of
these layers is due to fatigue cracking in the asphalt bound layers, stripping in the ATPB layer,
and softening of the aggregate base due to high moisture content in the aggregate base layer, and
high trafficking load levels.
Figure 22 presents elastic deflections as a function of the HVS load. These results
illustrate the non-linear response of the layers beneath the 250-mm pavement depth (all unbound
layers). The thick line in the figure delineates a linear elastic response. The importance of
properly characterizing the pavement materials is demonstrated by the results. In most
mechanistic-based analysis tools, to attain reasonable solutions, pavements are analyzed using
linear elastic theory, i.e., all layers are assumed to exhibit linear elastic response. For those
28
0
10
20
30
40
50
60
70
80
90
100
0 200000 400000 600000 800000 1000000 1200000 1400000
HVS Load Applications
Elas
tic D
efle
ctio
n C
ontr
ibut
ion,
Per
cent
0-250 mm250-430 mm430-850 mm>850 mm
40 kN Load 80 kN Load 100 kN Load
Layer
Figure 20. Contribution of pavement layers to surface elastic deflection based on 40-kN load, MDD 7.
0
100
200
300
400
500
600
700
800
900
0 200 400 600 800 1000 1200Elastic Deflections, microns
Pave
men
t Dep
th, m
m 10
236588
406098
457197
568356
638196
694154
779237
939343
1114295
HVS Repetitions
Figure 21. Profile of elastic deflections with depth.
29
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
30 40 50 60 70 80 90 100 110
Test Load, kN
Rat
io o
f Ela
stic
Def
lect
ions
to 4
0-kN
Ela
stic
Def
lect
ion
568356 Repetitions939343 Repetitions1197685 RepetitionsLinear Elastic
Figure 22. Elastic deflections versus test load level at pavement depth of 250 mm, MDD 7.
situations in which materials exhibit non-linear characteristics, the prediction of pavement
responses (stresses, strains, and deflections) is not correct. In turn, where these results are used to
estimate fatigue cracking, rutting, etc., predictions of pavement performance may be in error.
3.4.3 Back-calculated Moduli from In-depth Elastic Deflections
Moduli of the pavement layers were calculated from the in-depth elastic deflection basins
measured with the MDDs. Deflections were calculated using the Odemark-Boussinesq method
for a four layer system (asphalt bound layers, aggregate base layer, aggregate subbase, and
subgrade) and assuming a non-linear response of the subgrade at the positions where in-depth
deflections were measured with the MDDs (see Figure 2). Differences between measured and
calculated deflections were minimized by changing the moduli of the layers.(7)
30
Figure 23 summarizes the back-calculated moduli from the MDD data obtained for the
100-kN trafficking load. Back-calculated moduli for the 40- and 80-kN loads yielded what
appeared to be specious results and are not presented here.
Decrease in the modulus of the asphalt bound layer (including the ATPB) with HVS load
applications is presented in Figure 23. This reduction is due to fatigue cracking, as shown in
Figure 24, and by the deterioration of the ATPB due to stripping. A comparison of the moduli of
the asphalt bound layer (including the ATPB) back-calculated from MDD deflection with those
back-calculated from FWD deflections (see Section 3.6) indicates lower values than those back-
calculated from the MDDs. This difference is due at least in part, to the slower rate of loading at
which the tests were conducted. MDD deflections were measured under a slow moving wheel
load while FWD deflections were measured at a time of loading simulating a velocity of more
than 70 kph. Since the asphalt concrete modulus is dependent on time of loading, one would
expect a higher value at a shorter time of loading. Back-calculated moduli of the unbound
aggregate base are lower than those for the aggregate subbase and similar to the subgrade moduli
for the 100-kN load. This low modulus in the aggregate base likely was produced by an increase
in moisture content due to water infiltrating the base from the saturated ATPB. Comparison of
moduli determined from MDD deflections and those from the FWD at the end of HVS testing
suggests fair agreement among the two approaches. Back-calculated moduli of the aggregate
base and subbase (average of these two) from the MDDs are 14 percent lower than those back-
calculated from the FWD for the combined aggregate base and subbase. Back-calculated moduli
of the subgrade from MDDs were 40 percent lower than those from the FWD measurements.
Differences in moduli among the two approaches may result from differences in velocity, load
level, and wheel configuration at which the deflections were measured in the two tests. Change
31
1797
1167
818
590
330391 394
241297
41 4454 48 50 54
64
92 81
302
163 161 171
121
8167 62
44
62
10
100
1000
10000
- 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000
HVS Load Applications
Mod
ulus
(E),
MPa
E Asphalt Concrete + ATPBE Aggregate BaseE Aggregate SubbaseE Subgrade
Figure 23. Summary of back-calculated moduli from in-depth deflections.
0
500
1000
1500
2000
2500
0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000
HVS Load Applications
Def
lect
ion,
mic
rons 0 mm
213 mm396 mm625 mm1000 mm
MDD Depth
40 kN 80 kN 100 kN
Figure 24. In-depth elastic deflections for Section 502 (tested under dry conditions).
32
in the moduli of the asphalt concrete resulting from time of loading differences would change the
state of stresses in the unbound layers, which in turn affects the moduli of the unbound layers.
The HVS 100-kN dual wheel load versus that used in the FWD test (20 to 60 kN on a circular
plate), could also contribute to the differences.
3.4.4 Comparison with Previous Section Tested under Dry Conditions
Pavement Section 502 is similar to Section 543 and was previously tested under the HVS.
Differences between Section 502 and Section 543 and the traffic loading conditions for Section
502 are discussed in Section 3.2.3. In-depth MDD elastic deflections for the 40-kN test load are
presented in Figure 24. Elastic deflections at the beginning of the test were similar for both
sections, indicating similar responses under intact conditions, as shown in Figure 24. Significant
differences were observed under the 100-kN trafficking load. Elastic surface deflections level off
at about 500 microns for Section 502 and about 1000 microns for Section 543. The increase in
deflection in Section 543 indicates damage in the pavement structure and can be attributed to
failure of the ATPB layer.
3.5 Surface Cracking
Figures 25a, b, and c, respectively, show cracks detected during HVS trafficking after
about 951,000, 1.08 M, and 1.18 M load applications. The observed cracks were predominantly
transverse hairline cracks and were sometimes difficult to detect visually. The typical widest
crack widths were approximately 0.2 mm. These hairline cracks did not spall or increase
significantly in width during HVS trafficking. Lack of crack deterioration is attributed to
minimal amounts of dust (mineral particles) on the pavement surface. The HVS temperature
33
Section 543956471 Repetitions
0
100
200
300
400
500
600
050 -50
cm across section
cmal
ong
sect
ion
Section 5431075623 Repetitions
0
100
200
300
400
500
600
050 -50
cm across section
cmal
ong
sect
ion
Section 5431.18M Repetitions
0
100
200
300
400
500
600
050 -50
cm across section
cmal
ong
sect
ion
Figure 25. Surface crack schematics for Section 543.
34
control box, which limited daily expansion and contraction of the asphalt pavement, also likely
mitigated crack deterioration.
Inspection of the surface during HVS testing revealed deposits of fine silt accumulated on
the pavement surface along some of the cracks. The fine silt was periodically removed, but
reappeared after some trafficking. The fine silt was likely pumped from the aggregate base
through the cracks in the asphalt concrete layers to the surface by the rapid dissipation of water
pressure in the ATPB under the traffic loading of the HVS.
Figure 26 shows the extent of surface cracking along the test section at the completion of
HVS trafficking. More cracking was recorded between Stations 8 and 14 (400–700 cm) than
between Stations 2 and 8 (100–400 cm). These results are compatible with the deflections
measured with the FWD, which indicate higher FWD deflections where higher crack density was
measured.
Figure 27 shows the increase of crack density with HVS trafficking. A rapid increase in
crack density is observed after about 1,000,000 HVS load applications. These surface cracks
appeared and rapidly increased under trafficking at the 100-kN load. Referring again to Figures
16–18, it will be observed that after 1,000,000 HVS load applications, the elastic deflections
reached maximum values and remained at those levels for the duration of HVS trafficking.
3.6 FWD Test Results
FWD tests were conducted at intervals during HVS loading to assess changes in stiffness
of the various pavement layers. This section summarizes the results of deflection measurements
and moduli determined by back-calculation using the measured deflections.
35
45 0 -4525
75
125
175
225
275
325
375
425
475
525
575
3
4
5
6
7
8
9
10
11
12
13
Position acrosssection (cm)
Positionalong area of interst on section
(cm)
Stat
ion
543RF, 1.18M Repetitions: Crack Distribution
7.5-10
5-7.5
2.5-5
0-2.5
Cracking Density(m/m2)
Figure 26. Density of cracking on Section 543 after completion of HVS trafficking.
36
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 200000 400000 600000 800000 1000000 1200000 1400000HVS Repetitions
Tota
l Cra
ck L
engt
h, c
m
0.000.501.001.502.002.503.003.504.004.505.005.506.006.507.007.50
Crack D
ensity, m/m
2
Figure 27. Progression of crack density on Section 543.
3.6.1 Deflections
Figure 28 summarizes elastic deflections measured under the FWD load plate (D1)
normalized to a 40-kN load for the Stages 1, 2, and 4, as described earlier. As shown in Figure
28, D1 deflections showed no significant change before and during the water infiltration stages
(Stages 1 and 2). However, a significant increase in elastic deflections was observed after the
completion of HVS trafficking and during the forensic study (Stage 4). Table 9 summarizes these
results for each period of testing. In general, elastic deflections after HVS trafficking were about
3.9 times higher than those before HVS trafficking.
37
0
100
200
300
400
500
600
700
800
900
1000
0 2 4 6 8 10 12 14 16
Stations
Def
lect
ion,
mic
rons
May'99 (Stage 1: before water infiltration)July'99 (Stage 2: during water infiltration)Aug'99 (Stage 2: during water infiltration)Sept'99 (Stage 2: during water infiltration)Oct'99 (Stage 2: during water infiltration)Nov'99 (Stage 2: during water infiltration)May'00 (Stage 4: after HVS testing)
Figure 28. Summary of FWD elastic deflections.
Table 9 D1 Normalized Deflections for Section 543
Stage of Testing Date
Average Deflections (microns)
Standard Deviation (microns)
Average Temperature (ºC)
1 May 1999 127 6 24.0 July 1999 124 10 18.6 August 1999 130 12 21.0 September 1999 129 11 21.4 October 1999 126 11 23.0
2
November 1999 123 10 21.0 4 May 2000 490 37 18.0
3.6.2 Back-calculated Moduli from FWD Deflections
Moduli were back-calculated from FWD deflections obtained from three target loads (20,
40, and 60 kN) for a three layer system consisting of asphalt concrete (AC), aggregate base
(AB), and subgrade (SG). A three layer system was selected for simplicity. Back-calculated
38
moduli were determined using the program ELMOD 4.5 (Odemark-Boussinesq method
assuming a stress dependent subgrade modulus). For Section 543, back-calculations were
conducted for two cases to consider the ATPB layer since this layer typically has moduli values
between those of the asphalt bound layers and the aggregate base. In one case, the ATPB layer
was included as part of the asphalt bound layers and in the other case as part of the aggregate
base. Table 10 summarizes the layers and thicknesses for the two cases considered.
Table 10 Summary of Layers and Thicknesses Considered for Back-calculation Case 1 Case 2
Layer Thickness, mm
Layer Thickness, mm
AC = Asphalt Bound Layers + ATPB 247 AC = Asphalt Bound Layers 178
AB = Aggregate Base + Aggregate Subbase 268 AB = Aggregate Base +
Aggregate Subbase +ATPB 437
SG = Subgrade infinite SG = Subgrade infinite
Results of the back-calculation process are presented in Figures 29 through 32. Results
are presented for the three load levels and the two cases considered. Also presented in the figures
are results of the back-calculation process using the MDD deflections (Stage 3) for the 100-kN
dual wheel load.
Figure 29 shows the moduli of the asphalt concrete layers with and without the ATPB
layer. Moduli of the asphalt bound layers were higher than the moduli of the asphalt bound
layers including the ATPB for Stages 1 and 2. This indicates that the modulus of the ATPB is
significantly lower than that of the asphalt bound layers. Figure 30 shows a reverse effect for the
aggregate base when the ATPB is considered as part of this layer. The modulus of the aggregate
base and ATPB were considerably higher than the modulus of the aggregate base without the
ATPB, indicating that the modulus of the ATPB is higher than that of the aggregate base.
39
0
2500
5000
7500
10000
12500
15000
17500
20000
Mar-99 May-99 Jul-99 Aug-99 Oct-99 Dec-99 Jan-00 Mar-00 May-00
Time
Bac
kcal
cula
ted
Mod
ulus
, MPa
0
10
20
30
40
50
60
Asphalt C
oncrete Tempeture, ºC
AC+ATPB_1 AC+ATPB_2 AC+ATPB_3 AC_MDD 100kNAC_1 AC_2 AC_3 AC Temperature
Stage 1 Stage 2 Stage 3 Stage 4
Figure 29. Results of back-calculation, asphalt concrete layer with and without ATPB layer.
0
100
200
300
400
500
600
700
800
Mar-99 May-99 Jul-99 Aug-99 Oct-99 Dec-99 Jan-00 Mar-00 May-00
Time
Bac
kcal
cula
ted
Mod
ulus
, MPa AB_1
AB_2AB_3AB+ATPBAB+ATPBAB+ATPBAB_MDDASB_MDD
Stage 1 Stage 2 Stage 3 Stage 4
Figure 30. Results of back-calculation, aggregate base layer.
40
Using Odemark’s method of equivalent thickness, it is possible to estimate moduli of the
ATPB based on the results of the two cases considered. Figure 31 shows the results of this
approach. ATPB moduli for Stages 1 and 2 are within the same range as those obtained from
previous laboratory results. Figure 32 shows no significant difference in the modulus of the SG
layer for the two cases considered.
Relative to performance of the section, Figures 29 through 31 show the significant
deterioration of the bound layers and the aggregate base after HVS testing. Moduli of the asphalt
concrete layers were determined to be less than 1500 MPa at the conclusion of HVS testing.
Deterioration of the asphalt bound layers resulted from fatigue cracking as evidenced by the
surface cracks recorded during testing. Extracted cored indicated that cracks started from the
bottom lift of the asphalt concrete.
The rapid increase in surface cracking is likely the result of the significant reduction in
the stiffness of the ATPB and aggregate base, resulting in increased bending strains at the
underside of the asphalt concrete layers. Figure 31 demonstrates deterioration of the ATPB
produced by HVS trafficking. Estimated moduli of the ATPB after HVS trafficking were
between 200 and 400, a significant decrease from the initial measured moduli. As stated in
Section 3.2.2, the significant reduction in the stiffness of the ATPB was due to stripping resulting
from loading and the intrusion of fines from the aggregate base.
Damage in this aggregate base resulted from an increase in the moisture content and
increase in stress level due to the stiffness reduction in the ATPB layer (see Figure 30).
Figure 32 shows that the modulus of the subgrade experienced a significant reduction
(approximately 40%) after HVS testing. This reduction in modulus is the result of an increase in
the stress levels at the surface of the subgrade resulting from stiffness reductions of the upper
41
0
500
1000
1500
2000
2500
Mar-99 May-99 Jul-99 Aug-99 Oct-99 Dec-99 Jan-00 Mar-00 May-00
Time
Bac
kcal
cula
ted
Mod
ulus
, MPa
ATPB_1
ATPB_2
ATPB_3
Stage 1 Stage 2 Stage 3 Stage 4
Figure 31. Results of back-calculation, ATPB layer.
0
100
200
300
400
500
Mar-99 May-99 Jul-99 Aug-99 Oct-99 Dec-99 Jan-00 Mar-00 May-00
Time
Bac
kcal
cula
ted
Mod
ulus
, MPa
SG_1
SG_2
SG_3
SG_1(AB+AT
SG_2(AB+AT
SG_3(AB+AT
SG_MDD
Stage 1 Stage 2 Stage 3 Stage 4
Figure 32. Results of back-calculation, subgrade.
42
layers. Figure 33 includes average values for the layer moduli obtained from back-calculation for
the four stages of testing.
3.7 Ground Penetrating Radar (GPR) Surveys
GPR was conducted to estimate changes in the thickness of the asphalt concrete and
ATPB layers. Figure 34 shows the variation in thickness of the asphalt concrete and ATPB layer
during HVS trafficking. As shown in the figure, the GPR results indicate an average thickness of
the bound layers (asphalt rubber, asphalt concrete, and ATPB layers) of 253 mm. The GPR
estimate of the bound layer thickness is reasonable when compared with the as-built layer
thickness of 240 mm directly observed in the test pit. The GPR estimated an average reduction in
thickness for the bound layers across the whole trafficked section of about 25 mm at the end of
HVS trafficking, which is somewhat high compared to maximum surface rut depths measured
with the profilometer of approximately 23 mm.
Figure 35 shows the reduction of bound layer thickness across the section. In this figure,
the GPR data contours follow trends obtained using the profilometer. However, the GPR surveys
overestimate the reduction in layer thickness. According to the GPR surveys, the thickness
reductions were about twice the magnitude of those measured with the profilometer. Based on
the forensic activities, reduction in bound layer thickness was mostly due to the change in
characteristics of the ATPB layer in the trafficked areas. This overestimation by the GPR surveys
may be due to the contamination of the ATPB with fine material from the aggregate base, which
produces about the same GPR image as the aggregate base.
43
63
11201 13277
1327
281
16091767
343 304192 173
106
1
10
100
1000
10000
100000
Stage 1: nowater infiltration
Stage 2: waterinfiltration
Stage 3: HVStesting
Stage 4: afterHVS
Stage of Test Program
Laye
r Mod
ulus
, MP
aAC
ATPB
AB
SG
NO FWD TESTING
Figure 33. Average values of layer moduli obtained from back-calculation for the four stages of testing.
225
230
235
240
245
250
255
260
0 200000 400000 600000 800000 1000000 1200000 1400000
HVS Load Applications
Bou
nd L
ayer
Thi
ckne
ss, m
m
Figure 34. Variation in bound pavement layers as measured with Ground Penetrating Radar (GPR).
44
217k HVS load applications
514k HVS load applications
1,197k HVS load applications Figure 35. Reduction in bound layer thickness across Section 543 as measured with GPR.
45
3.8 Forensic Activities
Forensic activities included core extraction, dynamic cone penetrometer testing, and
trenching of the test section for direct observation of and data collection from the pavement
structure. The locations of forensic activities on Section 543 are shown in Figure 36.
3.8.1 Extracted Cores
Figures 37 and 38 show two extracted cores from Station 13. One core is from the
centerline of the trafficked area (Figures 37a and 37b) and the other is from about 1.4 m from the
centerline outside the trafficked area (Figure 36). Figure 37 shows the significant deterioration of
the asphalt concrete at the lift interfaces. This deterioration is likely due to lack of bonding
between the asphalt layers, which allowed for bending and movement along the interfaces.
During construction, a tack coat was applied between the AC and the asphalt rubber (ARHM-
CoresCores +DCPTestPit
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Station
0.0 0.5 1.0m
1000mm
500 mm
0 mm
-500mm
-1000mm
NTW
NTE
HW
HE
CLBW
BE
NTW
NTE
HW
HE
CLBW
BE
Non-Traffic West side
Non-Traffic East side
Hump West side
Hump East side
CenterlineBetween Non-Traffic & Centerline West side
Between Non-Traffic & Centerline East side
MDD
Figure 36. Locations of forensic activities on Section 543.
46
a. Asphalt Layers
47
Figure 37. Core extracted from centerline at Stati
b. Asphalt Treated Permeable Base on 13, Section 543.
Figure 38. Core extracted from outside the trafficked area at Station 13, Section 543.
48
GG) layers. A tack coat was not applied between the top and bottom AC lifts. The extracted core
showed no bonding between the layers. In addition, the ATPB could not be extracted as a solid
cohesive mass because the asphalt binder was completely stripped from the aggregate (see
Figure 37-b).
Figure 38 shows a core extracted from outside the trafficked area. Bonding was evident
through the layers with the ATPB layer still intact and attached, demonstrating that the ATPB
did not show any stripping in the non-trafficked areas.
Stripping in the ATPB resulted from traffic loading. The loading likely created high pore
water pressure in the ATPB since it was essentially in an undrained condition. The high pressure
resulted in infiltration of water into the binder and a reduction in its cohesion (stiffness). It also
probably caused water vapor to permeate the interfaces between the asphalt and aggregate. This
contributed to debonding and eventual stripping of the asphalt from the aggregate.
Figure 39 shows a major crack through the asphalt concrete layer. From earlier tests,
cracking was observed to have started at the bottom of the top layer of the asphalt concrete,
although in the sections which had been overlaid with either DGAC or ARHM-GG (RAC-G),
cracking eventually extended throughout the existing and overlay pavement layers.
Because of the reduced stiffness of the ATPB in the wet condition, it is hypothesized that
cracking started at the bottom of the DGAC and propagated to the pavement surface.
Figure 40 shows the significant contamination of the ATPB near its interface with the
aggregate base with fine silt material. It should be noted that the aggregate base had been primed
prior to the placing of the ATPB. The results of this test suggest that it would be desirable to
place a filter fabric on top of the untreated base prior to placing the ATPB to preclude upward
migration of fines when the drainage system is not effective.
49
Figure 39a. Front view.
50
Figure 39. Core extracted from centerline at Stati
Figure 39b. Opposite view.
on 9, Section 543.
Note: Sample is a slab shown upside-down (ATPB at top of image). Figure 40a. Fines-clogged ATPB slab from Section 543.
Note: Sample is shown upside-down (ATPB at top of image). Figure 40b. ATPB slab from another HVS test section demonstrating appearance of ATPB without fines infiltration. Figure 40. ATPB with and without fines infiltration.
51
3.8.2 Percolation Tests
Field percolation tests were conducted to measure the permeability of the ATPB and the
aggregate base layers at various HVS test locations. Figure 41 shows the location selected for the
field percolation tests. Table 11 summarizes the conditions in which the sections were tested
under the HVS.
Table 11 Description of Test Sections Used for Percolation Tests
Section ATPB Base Condition Traffic Load, kN
Asphalt Concrete Temperature (ºC)
Test Program Reference
505 Yes No water infiltration 40 50 9 512 Yes No water infiltration 40 50 9 518 No No Water infiltration 40,80, 100 20 10 543 Yes Water infiltration 40,80,100 20
Percolation test locations were classified based on the level of compaction added to the
ATPB and aggregate base due to trafficking. The following levels of compaction in the sublayers
(ATPB and aggregate base) were established:
• Level 1: No compaction in the sublayers due to HVS traffic. Percolation tests were
conducted outside of trafficked sections.
• Level 2: Compaction in the sublayers due to light HVS traffic. Percolation tests were
conducted on sections trafficked under a 40-kN load (Sections 505 and 512).
• Level 3: Compaction in the sublayers due to heavy HVS traffic. Percolation tests
were conducted on sections trafficked under 40-, 80-, and 100-kN loads (Sections 518
and 543).
Cores (150 mm) were extracted from the AC and ATPB layers in the trafficked and
untrafficked areas of the test section, as indicated in Figure 1. Because of the coring method, the
holes were approximately 160 mm in diameter. The AB was excavated by hand approximately
52
Sites of Field Percolation Tests Not to scale
Richmond Field Station, Bldg 280
Undrained Drained
Section 543Trafficked area for fatigue study
Section 512Trafficked area for rut study, super single
Section 505Trafficked area for rut study, dual
Section 518Trafficked area for fatigue study
Turnaround areasTurnaround areas with ramps
543
512505
518
N
10 32 4 65 87 9 10 11 1312 1514 16
10 32 4 65 87 9 10 11 1312 1514 16
1516 1314 12 1011 89 7 6 5 34 12 0
DG
ACA
RH
M-G
G
2%gradient
2%gradient
2%gradient
2%gradient
Figure 41. Sites of field percolation tests.
53
150 mm below the ATPB or AC, so that the bottom of the hole is approximately 50 mm above
the ASB. The holes were filled with water to the base of the ATPB or top of the AC, and allowed
to percolate through the AB for 24 hours. This delay period allowed the AB to saturate. After 24
hours, the water level was increased to the top of the AB and testing began. Care was taken to
ensure the water level was below the top of the AB during testing, thereby preventing water from
running between layer interfaces. The monitoring plan was scheduled to record the water level at
30 minute intervals for the first 4 to 6 hours and every hour thereafter. The percolation rate was
calculated as the drop in head per hour.
Figures 42 and 43 summarize the percolation rates for the aggregate base and ATPB
layers at the designated level of compaction. Figure 42 shows a contrast between aggregate base
beneath ATPB and without ATPB. The lines in Figure 42 show tendencies of the data using a
linear regression model. The contrast is probably due to the ATPB providing less intense stresses
on top of the aggregate base that would otherwise further compact the aggregate base layer, both
during construction and later during trafficking.
Figure 43 shows the dramatic difference between the high permeability of the ATPB
when this layer is intact and the permeability of the ATPB under the trafficked area of Section
543 (Test 543-5a). The permeability of the intact ATPB is about 1 cm/sec. The permeability of
the ATPB at the trafficked area in Section 543 was reduced about three orders of magnitude by
the fine material that infiltrated the ATPB layer from the aggregate base. This infiltration of fines
into the ATPB only occurred under the saturated base condition and when the section was
subjected to heavy loading from the HVS. Results from this are discussed earlier in this report.
54
505-15
505-4b
512-3
543-9
505-4a
512-4
543-13
543-5a
518-15
518-8
518-12
518-13
518-9
1.E-06
1.E-05
1.E-04
1.E-03
0 1 2 3Level of Compaction
Perm
eabi
lity
(cm
/sec
)
4
Percolation Rate with ATPBPercolation Rate without ATPB
Level of Compaction:1 - No Traffic Compaction: Minimal Post-Construction2 - Light Traffic Compaction: 40-kN Loading3 - Heavy Traffic Compaction: 40-, 80-, and 100-kN
Figure 42. Percolation tests on aggregate base.
543-5a
512-4
505-4a
543-5b505-15
512-3505-4b
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
1.E+01
0 1 2 3Level of Compaction
Perm
eabi
lity
(cm
/sec
)
4
Level of Compaction:1 - No Traffic Compaction: Minimal Post-Construction2 - Light Traffic Compaction: 40-kN Loading3 - Heavy Traffic Compaction: 40-, 80-, and 100-kN Loading
Figure 43. Percolation tests on ATPB.
55
3.8.3 Air-Void Content
Figure 44 summarizes air-void content data for all the bound layers after the completion
of HVS trafficking. Lower air-void contents were observed within the trafficked area compared
to outside the trafficked area for the ARHM-GG wearing course and the ATPB layer, with the
lowest air-void contents measured along the traffic centerline. A significant reduction in air-void
content (from 11–13 percent to about 9 percent) was observed for the ARHM-GG overlay as a
result of HVS trafficking. The high air-void contents observed in the ARHM-GG were
anticipated and reported in References (8) and (9). Figure 44 also shows the significant reduction
in air-void content in the ATPB layer. Air-void content of the ATPB may have been affected by
the fine silt contamination (Section 3.6.1). Small reductions in air-void content were observed in
the other asphalt concrete layers. Variations are attributed to construction variability.
13.1
11.4
9.1 9.610.9 11.0 11.0
5.06.4 6.2
7.1 7.2 6.7 6.3
2.94.4 4.6
3.6 3.5 2.9
21.0
18.317.6
24.0
0
5
10
15
20
25
30
-1500 -1000 -500 0 500 1000 1500
Pavement Location with Respect to Centerline, mm
Air-
Void
Con
tent
, per
cent
ARHMAC Top LiftAC Bottom LiftATPB
trafficked area
Figure 44. Average air-void contents across Section 543 after HVS trafficking.
56
Table 12 summarizes air-void contents in the bound layers in three distinct regions of the
test section: 1) trafficked area, which is the zone within the width of the HVS wheelpath
excluding the turnaround zones; 2) “hump areas,” which are the zones of uplifted material at the
edges of the trafficked area; and 3) non-trafficked areas, which are located outside of the HVS
wheelpath and beyond the hump areas.
Table 12 Summary of Air-Void Contents for Section 543
Average Air-Void Content in Region, percent Asphalt Layers Trafficked Area “Hump Areas” Non-Trafficked Areas ARHM Overlay 9.1 11.0 13.1
Asphalt Concrete, Top Lift 6.2 6.7 6.3
Asphalt Concrete, Bottom Lift 4.4 3.6 2.9
ATPB 17.6 19.2 21.0
3.8.4 Dynamic Cone Penetrometer (DCP) Data
Figures 45–47 summarize the results of Dynamic Cone Penetrometer (DCP) tests for
Section 543 performed inside and outside the trafficked area. Measurements were obtained
within a distance of ±1 m from the centerline. These figures show the amount of penetration
versus the number of blows applied. The slopes of these curves are proportional to the strengths
of the various pavement materials. A steeper slope (more rapid cone penetration) indicates a
layer with lower strength while a more gradual slope (lower penetration rate) indicates a stronger
material.
Table 13 summarizes DCP data for the aggregate base, aggregate subbase, and subgrade
after HVS trafficking. The data indicate that the aggregate subbase is slightly stronger than the
aggregate base.
57
0
200
400
600
800
1000
1200
0 50 100 150 200 250 300 350 400Blows
DC
P Pe
netr
atio
n, m
m
0.8 m West of Centerline0.25 m West of CenterlineCenterline0.25 m East of Centerline0.8 m East of Centerline
Figure 45. Dynamic Cone Penetrometer (DCP) tests at Station 5, Section 543.
0
200
400
600
800
1000
1200
0 50 100 150 200 250 300 350 400Blows
DC
P Pe
netr
atio
n, m
m
Centerline0.60 m East of Centerline1.0 m East of Centerline
Figure 46. Dynamic Cone Penetrometer (DCP) tests at Station 13, Section 543.
58
0
200
400
600
800
1000
1200
0 50 100 150 200 250 300 350 400Blows
DC
P Pe
netr
atio
n, m
m
0.5 m East of Centerline0.3 m East of Centerline0.3 m West of Centerline
Figure 47. Dynamic Cone Penetrometer (DCP) tests in trench, Section 543.
Table 13 Summary of DCP Penetration Rates, Section 543. Penetration Rate (mm/blow)
DCP Test Location Aggregate Base
Aggregate Subbase Subgrade
0.8 m W of Centerline 2.5 1.6 22.0 0.25 m W of Centerline 2.8 1.6 24.6 Centerline 2.9 2.4 19.8 0.25 m E of Centerline 2.3 0.4 24.6
Station 5
0.8 m E of Centerline 4.1 3.0 25.2 Centerline 2.6 2.6 20.7 0.60 m E of Centerline 3.0 1.2 19.6 Station 13 1.0 m E of Centerline 2.2 1.7 29.3 0.3 m W of Centerline 3.0 1.8 19.1 0.3 m E of Centerline 2.7 3.6 17.0 In Trench between
Stations 7 & 9 0.5 m E of Centerline 2.7 2.7 23.3
Average 2.8 2.1 22.3 Standard Deviation 0.5 0.9 3.5
59
3.8.5 Trench Data
After the completion of HVS trafficking, a transverse trench was excavated from Station
7 to Station 9 (see Figure 36) in order to enable direct observation of the pavement layers.
Measurements from the trench were used to estimate thickness and rutting in the pavement
layers. Figures 48 and 49 show profile data at the interface of each layer at the completion of
HVS trafficking. During the data collection process, it was difficult to establish the boundary
between the ATPB and the aggregate base due to the intrusion of fines from the aggregate base
into the ATPB. The profiles show that the subbase thickness varied considerably across Section
543. Differences in subbase thickness were anticipated and reported in previous CAL/APT
reports.(8, 10–12)
Figure 50 summarizes layer thickness inside and outside the trafficked area for the south
and north faces in the open pit. Layer thicknesses are average thicknesses across the transverse
distance at 500-mm intervals based on the transverse coordinate system shown in Figure 36.
The data indicate uniform thicknesses across the asphalt rubber and the two asphalt
concrete lifts. No significant difference in layer thickness was measured between centerline
thickness and thickness away from centerline. Therefore, significant rutting did not occur in
these layers. Standard deviations for the thicknesses of the asphalt rubber, asphalt concrete top
lift, and asphalt concrete bottom lift were 2, 3, and 4 mm, respectively.
Larger variability was found for the ATPB and aggregate base and subbase layers. Data
from outside the trafficked area (between –1000 and –500 and between 500 and 1000) indicate
thickness variability of the ATPB during construction. On average, the difference between the
two non-trafficked areas was between 12 to 17 mm. Standard deviation for the ATPB layer
across the two faces was 8 mm. Centerline thicknesses for the ATPB are thinner than the
thicknesses across the non-trafficked area between –1000 to –500 and thicker than the
60
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
0 100
200
300
400
450
550
Dep
th, m
m
ARHM Overlay
AC Top Layer
AC Bottom Layer
ATPB
AB
ASB
Subgrade
61
Figure 48. Layer profiles measured in trench, sout
650
750
850
950
1050
1150
1250
1350
1450
1550
1650
1750
1850
Transverse Distance, mm
Trafficked Area
h face of trench at Station 7, Section 543.
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
0 100
200
300
400
450
550
Dep
th, m
m
ARHM Overlay
AC Top Layer
AC Bottom Layer
ATPB
AB
ASB
Subgrade
62
Figure 49. Layer profiles measured in trench, nort
650
750
850
950
1050
1150
1250
1350
1450
1550
1650
1750
Transverse Distance, mm
Trafficked Area
h face of trench at Station 9, Section 543.
0
50
100
150
200
250
300
ARHM-GGOverlay
AC Top Lift AC BottomLift
ATPB AB ASB
Pavement Layer
Thic
knes
s, m
m
Centerline ± 50mm
-1000mm to -500mm-500mm to 0mm
0mm to 500mm500mm to 1000mmAverage
Figure 50. Summary of layer thicknesses across transverse distance of trench.
thicknesses across the non-trafficked area between 500 to 1000 mm. Therefore, no clear
conclusion can be drawn regarding rutting in this layer based on the data from the test pit.
However, the reduced thickness of the ATPB and the amount of fine intrusion from the
aggregate base along the centerline of the section ±100 mm suggest deformations in the ATPB
and aggregate base layers at the interface.
Thickness measurements for the aggregate base obtained at the centerline are lower than
those obtained outside the trafficked areas. Differences were 6 and 13 mm for the north face and
south face, respectively. Standard deviation for the aggregate base layer across the two faces was
8 mm. The data also show the thickness variation of the aggregate subbase across the transverse
distance of the section. The standard deviation for the aggregate subbase thickness was 27 mm.
63
Based on the average thickness measurements obtained for the pavement layers, rutting
occurred primarily in the area near the interface of the ATPB and the aggregate base. However,
due to rough interfaces between the ATPB and aggregate base, between the aggregate base and
the aggregate subbase, and between the aggregate subbase and the subgrade, accurate estimates
of the thicknesses of the ATPB, aggregate base, and aggregate subbase and rutting estimates in
each of these layers were difficult to determine.
64
4.0 PERFORMANCE EVALUATION AND MECHANISTIC ANALYSIS
By the end of HVS testing, Section 543 had been subjected to 236.6 ×103 applications of
the 40-kN load, 331.8 × 103 applications of the 80-kN load, and 629.3 × 103 applications of the
100-kN load. The number of applied loads was converted to equivalent 80-kN single axle loads
(ESALs) using a load equivalency exponent of 4.2 according to Caltrans procedure, resulting in a
total of 35.8 million ESALs. According to the failure criterion established for the project, the
number of ESALs to failure for surface rutting was 18.7 million while that for surface cracking
(2.5 m/m2) was 35.8 million. It is evident that Section 543 failed by surface rutting before
reaching the surface cracking criterion.
As discussed in Chapter 3, Section 543 failed because of stripping and intrusion of fines
into the ATPB layer. In the following paragraphs, estimates of pavement responses at the
beginning and end of HVS testing, based on the layer moduli obtained during the back-
calculation process, are provided together with a discussion of the impact of the response on the
performance of the pavement section. Three cases were considered for the analyses. For Cases I
and II, the stiffness characteristics of the pavement structure were defined by the FWD back-
calculation process (see Section 3.4.3). For Case III, the ATPB layer was replaced by an
equivalent layer of dense graded asphalt concrete. While direct comparison of Cases I and II with
Case III is not possible because the expected mode of failure of the pavement structure for Case
III is fatigue cracking rather than rutting failure due to stiffness degradation in the ATPB, Case
III has been included to question the necessity of using an ATPB layer.
4.1 Pavement Responses
Pavement responses considered for the analyses are tensile strain at the bottom of the
asphalt concrete (ATPB is not included), and vertical compressive stresses at the top of the
65
ATPB, aggregate base, and subgrade layers. Layer moduli values were obtained from FWD
testing by back-calculation. For Case III, the 75-mm ATPB layer was replaced with an
equivalent thickness of DGAC obtained using the Odemark method. For a modulus of 1,600
MPa for the intact ATPB and 10,000 MPa for the intact DGAC layer, the equivalent thickness
would be about 40 mm of DGAC.2 Materials, thicknesses, and moduli for the pavement
structures used for the three cases are summarized in Table 13.
Table 13 Summary of Pavement Structures for Analysis
Case Layer Thickness, mm
Modulus before HVS Trafficking, MPa
Modulus after HVS testing, MPa
AC + ATPB 247 8,500 950 AB 368 300 70 Case I SG 170 100 AC 178 12,000 1,900 ATPB + AB 477 470 140 Case II SG 160 120 AC 218 10,000 - AB 268 300 - Case III SG 170 -
Table 14 summarizes calculated pavement responses under a 40-kN load for the three
cases considered. Tensile strains at the bottom of the AC layer have been used to estimate the
fatigue performance of the pavement structures.
Table 14 Summary of Pavement Responses under 40-kN Load
Case
Before or After HVS Trafficking
Tensile Strain, Bottom of AC
Vertical Stress, Top of ATPB, kPa
Vertical Stress, Top of AB, kPa
Vertical Stress, Top of SG, kPa
Before - - 32 19 Case I After - - 50 46 Before 46 63 - 17 Case II After 222 91 - 37 Before 42 - 36 21 Case III After - - - -
2 mmEE
hhDGAC
ATPBATPBDGAC 40
10000160075 33 ≅==
66
4.2 Fatigue Analysis
The fatigue analysis and design system used in this report is shown in Figure 51.(10) For
the analysis, a design reliability of 50 percent, corresponding to a reliability multiplier of 1.0,
was assumed. The laboratory fatigue life equation for the mixes used in the simulation is:
εln14248.4106663.09295.21ln −−−= AVN
where N is the number of repetitions to failure obtained from laboratory fatigue beam tests using
the controlled strain mode of bending, AV is the percent air-void content of the mix, and ε is the
tensile strain at the bottom of the asphalt concrete. The shift factor (SF) equation, which has been
calibrated using pavement designed according to the Caltrans procedure, is:
3759.15101833.3 −−×= εSF
The temperature conversion factor (TCF) equation used for this simulation3 is:
891.2)ln(754.1 −= dTCF
where d is the thickness of the asphalt concrete layer in centimeters.
The results of the simulation process are summarized in Table 15 for the information
obtained from the three cases. Cases I and II represent the same pavement structure using
pavement responses measured at different locations.
Table 15 Summary of Calculation of ESALs Using UCB Fatigue Analysis System Case Microstrain Air-
Void Content, %
ln N N, × 106
Shift Factor
AC thickness (tAC), cm
Temperature Conversion Factor (TCF)
M ESALs, millions
I-II 46 4.6 18.95 170.5 29.5 17.8 2.16 1 2,333 III 42 4.6 19.33 248.6 33.5 21.8 2.51 1 3,310
3 The equation for TCF is for the region along the California Coast in the vicinity of Santa Barbara. It tends to provide more conservative estimates of traffic than other areas of California
67
allowable ESAL
Laboratory fatigue life (N): N = a ε b
Temperature Conversion Factor (TCF): TCF = 1.754 ln(d) - 2.891d is AC thickness in cmTCF has been calculated for Californiadesert, mountain, coastal environments
tensile strain (ε) at bottom of asphalt concrete layer from laelastic theory calculation of response of pavement to 80 kNsingle axle, dual tires
Asphalt Concrete
Base
Subgrade
Pavement Structure Represeby Multi-Layer Elastic Solid
Layered ElasticAnalysis
* stiffness or elastic modulus, thicknessand poisson ratio required for each layer
68
Figure 51. Methodology followed in the fatigue ana
s = N * SF TCF * M
Shift factor (SF):SF = 2.7639x10-5 ε -1.3586
calibrated against Caltransdesign procedure, accounts fortraffic wander, crack propagation
Reliability Multiplier (M):
yer
a, b from laboratory fatigue beam tests
Variance of trafficdemand estimate
(ESALs)
Variance oflaboratory fatigue
test results
mix stiffness from laboratory fatigue beam tests
nted * Laboratory Beam Fatigue Test of Asphalt Concrete
( ) ( )ESALsNze lnvarlnvar M +=
lysis system to determine ESALs.
The simulations suggest that if fatigue failure were the distress mode, the sections would
sustain at least 2,000 × 106 ESALs (for a reliability level of 50 percent). However, the results of
HVS testing of Section 543 indicated that the fatigue performance of this type of section can be
severely affected if the ATPB strips. While little fatigue cracking was observed in the section up
to the point that the ATPB started to strip, rapid development of surface cracking was observed
after this point. It should be noted that if the ATPB is designed in such a way that it does not
strip, then the performance of this type of section would be excellent. To improve ATPB
performance requires that an improved mix design procedure be developed for the ATPB and
that a filter layer (fabric) should be used at the interface between the ATPB and untreated base to
prevent the intrusion of fines into the ATPB.
The simulations also suggest that the pavement structure for Case III will outperform the
pavement structure of Cases I and II. Replacing the ATPB with an equivalent thickness of
asphalt concrete provides an improved fatigue life and reduction in stresses in the unbound
layers.
4.3 Economic Analysis
Selecting a pavement structure with the ATPB layer or one with an additional thickness
of AC would depend on which pavement structure is the most cost-effective. Based on Caltrans
construction data, between 1994 and 1998, the cost of ATPB averaged $56 per cubic meter for
an average annual production of 81 thousand cubic meters. For the same period, the cost of Type
A dense graded asphalt concrete was $35 per tonne, or about $83 per cubic meter, about 1.5
times the cost of ATPB. The cost savings achieved by replacing the 75-mm ATPB layer with a
40-mm DGAC lift would be $880 per kilometer per meter width of pavement. This cost savings
plus the benefit of eliminating a potential stripping layer makes the 40-mm DGAC lift attractive.
69
However, it must be emphasized that to achieve this cost savings while maintaining
performance, the asphalt concrete lifts must be properly compacted.
4.4 Rutting Analysis
Rutting failure in Section 543 has been attributed to stripping of the ATPB layer. As
indicated in the previous section, the failure in the ATPB and subsequent fatigue cracking in the
asphalt concrete layers increased stresses at the top of the aggregate base. In addition, because of
intrusion of fines from the aggregate base into the ATPB, the permeability of the ATPB was
reduced, causing water to be retained in the ATPB layer. The retention of water allowed the
aggregate base to increase in water content, thus making it more susceptible to rutting. While
criteria for unbound layer rutting are still under development because of the complex state of
stresses in the unbound materials, a relatively straightforward method is to determine the ratio of
the applied stress to the stress required to cause shear failure in the material.
Based on preliminary laboratory triaxial compression test data for untreated granular
materials, stresses to cause failure are in the range of 250 to 800 kPa for confining pressures
ranging from 0 to 105 kPa. These results are based on tests conducted on specimens compacted
to a relative compaction of 95 percent and optimum moisture content based on California Test
Method 216. Since computed vertical stresses were in the range of 30 to 50 kPa for a 40-kN load
(see Table 14), stress ratios of applied stress to stress to failure were in the range of 0.03 to 0.2.
Ratios below about 0.8 are not expected to cause failure in untreated aggregate base.(13)
The significant amount of rutting in the aggregate base (10 mm at the end of HVS
trafficking) was due to the increased stress produced by the 100-kN load. Computed vertical
stresses on top of the aggregate base were in the range of 70 to 120 kPa before HVS trafficking.
Due to the progressive failure of the ATPB, it is believed that the lack of confining pressure, loss
70
of fine particles, and increase in moisture content significantly reduced the required stress to
failure in the aggregate base, thereby increasing the stress ratios and making the aggregate base
susceptible to rutting.
A similar approach was used to assess subgrade performance. Estimated stresses to
failure under unconfined conditions were approximately 200 kPa for a moisture content of 16
percent. Compute stresses were in the range of 20 to 50 kPa. Therefore, stress ratios were in the
range of 0.1 to 0.25. For cohesive subgrades, rutting failure is not expected to occur if stress
ratios are below 0.5.(13)
Recommendations to improve the rutting performance of unbound materials include
increasing the in-place density of these materials. The laboratory tests conducted on the unbound
materials used in the Goal 5 project substantiate the improvement in performance due to increase
in density.(14)
71
72
5.0 SUMMARY AND CONCLUSIONS
This report has presented the results of the accelerated pavement testing program for
Section 543 of the Goal 5 research project. This section is a flexible pavement which includes 75
mm of an ATPB layer between the asphalt concrete and the aggregate base layers. Included in
the report are the analyses of the four stages of testing considered in the Goal 5 test plan. Test
program results provide recommendations relative to the use of ATPB layers in flexible
pavement structures.
The test program for Section 543 started in May 1999 and was completed in March 2000.
Under HVS trafficking, a total of 1.197 × 106 repetitions were applied on the section, including
236 × 103 repetitions of the 40-kN dual wheel load, 332 × 103 of the 80-kN load, and 629 × 103
of the 100-kN load. Trafficking was performed under “wet base” conditions accomplished using
a water infiltration system. The wet conditions were considered representative of infiltration
associated with precipitation during an average peak week of 51.3 mm, typical for Eureka, CA
(California North Coast climate region).
Initial rutting and deflection data collected during HVS trafficking under the 40- and 80-
kN loads indicated that the pavement performed about the same as similar sections tested under
dry conditions. During the 100-kN loading, section performance deteriorated significantly as
evidenced by an increased rate of surface rutting and fatigue cracking. The section initially failed
according to a surface rutting criterion of 12.5 mm. Associated forensic data attributed the failure
to a reduction in stiffness in the ATPB layer, which resulted from stripping of the ATPB caused
by its saturated state and traffic loading as well as intrusion of fines from the untreated aggregate
base.
In-depth elastic deflection and permanent deformation measurements obtained during
traffic loading indicated a problem in the area above and below the interface of the ATPB and
73
the aggregate base. Subsequent sampling of materials during the forensic study indicated that the
ATPB exhibited significant stripping in some locations in the trafficked areas. At other locations,
air-void contents of the ATPB in the trafficked area were less than obtained from cores extracted
in the non-trafficked areas.
Percolation tests conducted on the trafficked areas and outside the trafficked areas
indicated a significant reduction in the desirably high initial permeability of the ATPB layer.
Examination of the open trench during the forensic study confirmed that fine particles from the
aggregate base had infiltrated and clogged the ATPB layer.
FWD testing conducted before and after HVS trafficking indicated that the ATPB
exhibited a significant loss in stiffness with the stiffness approaching that of the unbound
material. Due to the reduced stiffness of the ATPB, significant rutting occurred on top of the
aggregate base due to the lack of confinement resulting from the reduced stiffness of the ATPB
together with cracking of the asphalt concrete layers. The increase in rutting in the aggregate
base resulted from an increase in the bending stresses at the bottom of the asphalt concrete layers
producing an accelerated increase in surface cracking. The limited number of cores extracted
along the centerline of HVS trafficking indicated that cracking occurred in all asphalt concrete
lifts. These cores also indicated that no bond existed between asphalt concrete lifts.
5.1 Conclusions
From the results of the tests on Section 543 and associated response and performance
analyses, the following conclusions are drawn:
1. The stiffness of the ATPB was significantly reduced due to stripping, which
resulted from high stresses (both pore water and interparticle) and the presence of
water during traffic loading. Moreover, the prime coat on the untreated aggregate
74
base was ineffective in preventing the migration of fines from the aggregate base
into the ATPB.
2. Intrusion of fines from the aggregate base into the ATPB reduced the permeability
of the ATPB layer by about three orders of magnitude (from about 1 cm/s to 10-3
cm/s). This filling of voids in the ATPB resulted in a system in which additional
water was retained in the ATPB; the higher degree of saturation resulted in
reduced stiffness which led to accelerated deterioration of its load carrying
capability.
3. Section 543 failed by rutting due to reduced stiffness in both the ATPB and the
aggregate base. The rapid increase in rutting also contributed to acceleration in
surface cracking because of the significant increase in tensile strains in the asphalt
concrete layer.
5.2 Recommendations
1. While the ATPB can provide an increased structural capacity to asphalt concrete
pavement when loaded under dry conditions, it is imperative that the ATPB be
designed to be resistant to stripping and consequent loss in stiffness, should water
be retained in the ATPB due to a blocked drainage system.
2. Improved ATPB performance necessitates an improved mix design procedure,
improved drainage design, and improved construction and maintenance
procedures. In terms of mix design, use of an increased asphalt content and the
use of a modified binder may reduce the susceptibility of the ATPB to water
damage.
75
3. Improved drainage design requires the use of a filter layer (e.g., filter fabric)
adjacent to the ATPB to minimize the intrusion of fines from the untreated base
since the prime coat is not sufficient. In addition, edge and transverse drains must
be maintained to prevent becoming clogged with fines.
4. The use of ATPB to intercept water entering through the pavement surface should
be reconsidered. As an alternative, if the permeability of the asphalt concrete is
reduced by means of improved compaction, a sufficient thickness of asphalt
concrete is provided to mitigate the potential for load associated cracking, and the
degree of compaction of the aggregate base is increased, the need for the ATPB
between the asphalt concrete and the aggregate base could be eliminated.
76
6.0 REFERENCES
1. Harvey, J., M.O. Bejarano, A. Ali. Test Plan for Goal 5: Performance of Drained and Undrained Flexible Pavement Structures under Wet Conditions. Pavement Research Center, Cal/APT Program, Institute of Transportation Studies, University of California, Berkeley.
2. Harvey, J., Chong, A., Roesler, J. Climate Regions for Mechanistic-Empirical Pavement Design in California and Expected Effects on Performance. Draft report prepared for California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. June 2000.
3. Harvey, J., B-W Tsai, F. Long, and D. Hung. CAL/APT Program: Asphalt Treated Permeable Base (ATPB), Laboratory Tests, Performance Predictions and Evaluation of Caltrans and Other Agencies Experience. Report prepared for the California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. June 1999.
4. Bejarano, M., J. T. Harvey, A. Ali, M. Russo, D. Mahama, D. Hung, and P. Preedonant. Performance of Drained and Undrained Flexible Pavement Structures under Wet Conditions: Test Data from Accelerated Pavement Test Section 544-Undrained. Report in process for the California Department of Transportation. Pavement Research Center, Institute of Transportation Studies, University of California, Berkeley. Expected date of publication: June 2004.
5. Bejarano, M., J. T. Harvey, A. Ali, M. Russo, D. Mahama, D. Hung, and P. Preedonant. Performance of Drained and Undrained Flexible Pavement Structures under Wet Conditions: Test Data from Accelerated Pavement Test Section 545-Undrained. Report in process for the California Department of Transportation. Pavement Research Center, Institute of Transportation Studies, University of California, Berkeley. Expected date of publication: June 2004.
6. Harvey, J. T., L. du Plessis, F. Long, S. Shatnawi, C. Scheffy, B-W. Tsai, I. Guada, D. Hung, N. Coetzee, M. Reimer, and C. L. Monismith. Initial CAL/APT Program: Site Information, Test Pavement Construction, Pavement Materials Characterizations, Initial CAL/APT Test Results, and Performance Estimates. Report prepared for the California Department of Transportation. Report No. RTA-65W485-3. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley, June 1996, 305 pp.
7. Ullidtz, P. Pavement Analysis. Elsevier Publishing Company, 1987.
8. Harvey, J., J. Prozzi, J. Deacon, D. Hung, I. Guada, L. du Plessis, F. Long, and C. Scheffy. CAL/APT Program: Test Results from Accelerated Pavement Test on Pavement Structure Containing Aggregate Base (AB)--Section 501RF. Report prepared for the California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. Draft submitted September 1997. Final submitted April 1999.
77
78
9. Harvey, J., M. O. Bejarano, A. Fantoni, A. Heath, and H. C. Shin. Performance of Caltrans Asphalt Concrete and Asphalt-Rubber Hot Mix Overlays at Moderate Temperatures – Accelerated Pavement Testing Evaluation. Draft report submitted to California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. July 2000.
10. Harvey, John T., Louw du Plessis, Fenella Long, John A. Deacon, Irwin Guada, David Hung, Clark Scheffy. CAL/APT Program: Test Results from Accelerated Pavement Test on Pavement Structure Containing Asphalt Treated Permeable Base (ATPB) Section 500RF. Report No. RTA-65W485-3. Report Prepared for California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. June 1997.
11. Harvey, J., I. Guada, C. Scheffy, L. Louw, J Prozzi, and D. Hung. CAL/APT Program: Test Results from Accelerated Pavement Test on Pavement Structure Containing Asphalt Treated Permeable Base--Section 502CT. Draft report submitted to California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. February 1998.
12. Harvey, J., D. Hung, J. Prozzi, L. Louw, C. Scheffy, and I. Guada. CAL/APT Program: Test Results from Accelerated Pavement Test on Pavement Structure Containing Untreated Aggregate Base--Section 503RF. Draft report submitted to California Department of Transportation. Pavement Research Center, CAL/APT Program, Institute of Transportation Studies, University of California, Berkeley. December 1997.
13. Bejarano, M. Subgrade Soil Evaluation for the Design of Airport Flexible Pavements. Ph.D. thesis, University of Illinois, Urbana-Champaign. 1999.
14. Heath, A. C. Modelling Unsaturated Granular Pavement Materials using Bounding Surface Plasticity. Ph.D. thesis, University of California, Berkeley. 2002.